British Journal of Ophthalmology, 1983, 67, 244-248

Enolase isoenzymes in uveal melanomas-a possible parameter of malignancyJANICE A. ROYDS,' I. G. RENNIE,2 M. A. PARSONS,3 W. R. TIMPERLEY,4 AND

C. B. TAYLOR

From the Departments of 'Biochemistry, University of Sheffield; 2Ophthalmology, Royal HallamshireHospital, Sheffield; 3Histopathology, University ofSheffield; and 4Neuropathology, Royal Hallamshire Hospital

SUMMARY Seven uveal melanomas were stained for y enolase by an immunoperoxidase PAP(peroxidase-antiperoxidase) technique, and biochemical assays were carried out on tissue homo-genates. A correlation between the biochemical assay and the immunoperoxidase staining wasdemonstrated. Two tumours with the highest biochemical assays showed positive staining for theenzyme, and 2 tumours with the lowest levels showed no appreciable staining. The highest level ofenolase was present in a tumour which both- clinically and histologically appeared to be benign, andthe lowest level occurred in a mixed cell tumour which was large in size; the low level presumablyrelated to its relatively fast rate of growth. Estimation of y enolase activity in ocular melanomasmay provide an accurate quantitative method for assessing the malignant potential of thesetumours.

Recent changes in attitudes to the treatment of uvealmelanomas have heightened the need for an accuratemethod of histological assessment of their prognosis.Most pathologists use a modification of the classifica-tion devised by Callender in 1931.' This systemconsists of 4 histological groups-spindle cell A,spindle cell B, mixed cell, and epithelioid cell. Thedetermination of the cell type, based on this classifica-tion, appears to be the most important parameter inassessing prognosis.2` Tumours composed in part(mixed cell) or completely of epithelioid cells have arelatively poor prognosis. Unfortunately it has beenshown that there is often a considerable disparity inthe histological grading of these tumours when theyare examined by different pathologists.

In a study performed by Gass6 15 pathologists wereasked to grade 15 uveal melanomas independently. Inonly 4 cases was there unanimous agreement on theappropriate histological grade. The problem ofaccurate classification most commonly arises from thedifficulty in identifying epithelioid elements withinthe tumour. Furthermore uveal melanomas oftenshow a marked difference in cell type from one area toanother, and the pathologist may fail to classify the

Correspondence to Dr W. R. Timperley, Department of Neuro-pathology, Royal Hallamshire Hospital, Glossop Road, SheffieldS10 2JF.

tumour correctly because the section examined maynot be representative of the tumour as a whole.

Enolase (E.C.4.2.1.11) is a dimeric cytoplasmicenzyme which catalyses the interconversion of2-phospho-d-glycerate to phosphoenolpyruvate in theglycolytic pathway. It has been shown to contain 3distinct subunits designated a, /3, and y, and 5isoenzyme types have been demonstrated.7" Theseinclude the 3 homodimers aa, 3(13, and yy and 2hybrids a/3 and ary.a Enolase is the commonest form in most adult

tissues and is probably the sole isoenzyme present inearly fetal tissue. '0 Studies on monkey and rodenttissues indicate that y enolase is confined to neuronesand cells of the amine precursor uptake anddecarboxylation (APUD) system, 1213 and our ownpersonal observations indicate that this is also true inthe human. Melanocytes, a component of the APUDsystem'4 have been shown to contain y enolase. "Royds et al.'6 suggested that with increasing de-

differentiation of cutaneous melanocytic lesions thereis a reversion to the fetal enzyme pattern and a.com-parative loss of y enolase from the neoplastic tissue.The presence of epithelioid elements appears to havean adverse effect on the prognosis of uvealmelanomas. Thus one might anticipate a reduction inthe level of y enolase in tumours containing theseelements, and estimation of the amount of -y enolase

244

on February 1, 2020 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.67.4.244 on 1 A

pril 1983. Dow

nloaded from

Enolase isoenzymes in uveal melanomas-a possible parameter of malignancy

within uveal melanomas may be useful in assessingprognosis.

This paper presents the results of the localisationand assay of y enolase in 7 uveal melanomas by meansof the immunoperoxidase PAP technique onformalin-fixed paraffin sections and direct bio-chemical assay of tissue homogenates.

Patients and methods

PATIENTSSpecimens were obtained from 7 patients who had aneye enucleated for a primary choroidal melanoma.The age of the patients in the study ranged from 40 to88 years (mean 64-7 years). Patient 1 had beenobserved for 21/2 years prior to the enucleation. Thetumour was relatively small, and little growth hadbeen observed during that period. In view of the smallsize of the tumour and the patient's age enucleationwas not proposed at the time of diagnosis. The eyewas eventually enucleated because a painfulthrombotic glaucoma had developed. Patient 3 hadbeen observed for one year prior to enucleation. The

tumour appeared to have arisen from a choroidalnaevus. The eye removed from patient 7 had beendiagnosed as having melanosis oculi with externalpigmentation, heterochromia iridis, and a deeplypigmented choroid. The tumour in this eye was large,occupying at least half the globe.

METHODSImmediately after enucleation the globe was opened,and the tumour was identified and biopsied, wherepossible from more than one site. The remainder ofthe tumour was left in situ in the globe and fixed informol saline. The biopsy material removed wasrapidly frozen and stored at -80°C until biochemicalassay was performed.

Histological methods. After fixation in formolsaline for a period of 24 to 48 hours sections from theglobe containing the tumour were processed intoparaffin wax. Standard 5 gm sections were cut andstained with haematoxylin and eosin. Histologicalgrading by the modified Callender classification wasperformed independently by one of us (I.G.R.).

Immunohistochemistry. Paraffin sections were

Fig. 1 Spindle celled melanomastained for y enolase. (x 800).

245

on February 1, 2020 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.67.4.244 on 1 A

pril 1983. Dow

nloaded from

Janice A. Royds. L G. Rennie, M. A. Parsons, W. R. Timperley, and C. B. Taylor

Table I Gamma enolase in various types ofuveal melanoma

Series Age Sex Site Cell type PAP y y ynumber

pmg/g ,Lg/mg protein

1 88 F Choroid Spindle B (with areas spindle A) 1-5 + patchy 46-6 0-632 59 F Choroid Spindle B I + 40 9 0 603 40 M Choroid Mixed, predominantly epithelioid + 21 -5 0-224 70 F Choroid Mixed, predominantly epithelioid + 10 0-205 67 M Choroid Mixed, predominantly spindle + 8-96 0-156 56 M Choroid Mixed, predominantly spindle -ve 4-0 0107 73 F Choroid Mixed, no predominance of cell type -ve 1 86 0 04

stained for y subunits of enolase by the peroxidase-antiperoxidase (PAP) technique. Sections (5 ,um)were stained by the peroxidase-antiperoxidase tech-nique using anti-y enolase sera. Normal rabbit serumand adsorbed antisera were used as controls. Theintensity of the staining for y enolase was gradedindependently by one of us (W.R.T.).

Biochemical assay-treatment of tissue samples.Tissue samples were weighed, homogenised in 2 ml of50 mM tris buffer, pH 7*1, containing 4 mM MgC12,and then centrifuged at 40000 g for 30 minutes at 4°C.The resulting supernatant fluids were assayed in

Fig. 2 Spindle celled melanoma,negative control. Note thefew cellscontaining granules ofmelaninpigment in the bottom right-handcorner. (x 800).

duplicate for y immunoreacting enolase by radio-immunoassay.Radioimmunoassay. The radiolabelled yy enolase

was prepared by the method of Bolton and Hunter'7using N-succinimidyl-3,4 hydroxy-5 ('25I) iodophenylpropionate (Amersham International Ltd..Amersham, Bucks HP7 9LL).The primary antiserum used was produced in

rabbits. 8 The separation of the bound and freeantigen in the assay was accomplished by precipita-tion with goat anti-rabbit IgG as a secondary antibody(Miles Laboratories Ltd., Slough, Bucks). The tubes

246

on February 1, 2020 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.67.4.244 on 1 A

pril 1983. Dow

nloaded from

Enolase isoenzymes in uveal melanomas-a possible parameter of malignancy

were then centrifuged at 2000 g for 30 minutes at 4°Cin an MSE6L. The supematants were removed byaspiration and the pellets counted in an Intertech-nique Well-type y counter. The y immunoreactivityin the melanomas became diluted in parallel with thepurified yy enolase usedc to generate the standardcurve, providing evidence that the 2 cross-reactimmunologically.

Results (Table 1)

Histological grading. Five of the tumours were gradedas mixed cell tumours, the other 2 being spindle cellB. In one case (patient 1) large areas of spindle cell Acells were present within the tumour. However, inkeeping with the guidelines laid down by McLean etal.19 on the classification of spindle cell A tumours thistumour was classified as spindle cell B.

Identification of y enolase in paraffin sections ofuveal melanomas by the immunoperoxidase tech-nique. The 2 spindle cell tumours with the highest yenolase levels both stained positively for y enolase(Figs. 1, 2). The strongest stainingwas found in patient1, whose tumour was classified as spindle cell B con-taining spindle cell A elements. Of the 5 mixed celltumours 3 were found to have equivocal staining for yenolase, and in 2 the staining for y enolase wasabsent.

Estimation of y enolase by biochemical assay. yEnolase was found in all 7 tumours studied (Table 1).However, a marked variation in the levels presentwas found. The highest levels of y enolase, 46&6 ,ug/gand 40 0 gg/g were found in 2 of the spindle cell Btumours. All 5 of the mixed cell tumours had lowerlevels of the enzyme than the spindle celled tumours.

Discussion

Our results confirm the presence of y enolase withinmelanocytic tumours of the uveal tract. Although theseries is small, the more benign spindle cell tumoursappear to have a higher level of y enolase than thosecontaining epithelioid cell clements. The highest levelof enolase was present in a tumour which both clinic-ally and histologically appeared the most benign. Thelowest enolase level occurred in a mixed cell tumour(patient 7) which was large in size; its size may reflecta relatively fast growth rate. It has been establishedthat large melanomas of the uveal tract carry arelatively poor prognosis.A correlation between the biochemical assay and

the immunoperoxidase staining was demonstrated.The 2 tumours with the highest biochemical assaysboth stained positively for the isoenzyme. The 2tumours with the lowest levels of y enolase showed noappreciable staining.

Both methods of y enolase estimation may be ofvalue in estimating the malignant potential of a uvealmelanoma. The biochemical assay, although requir-ing the acquisition of fresh material in order to per-form it, has distinct advantages. Firstly, unlike theimmunoperoxidase method, it is quantitative ratherthan subjective. Secondly, the larger amount ofmaterial analysed by this technique renders it lesssusceptible to sampling error. Obviously further workon y enolase levels in uveal melanomas is necessary.This should include a prospective trial where enolaselevels within uveal melanomas could be comparedwith long-term survival. Nevertheless we believe thaty enolase estimation may provide an accurate quanti-tative method for assessing the malignant potential ofuveal melanomas.

The authors thank Messrs A. J. Dark. R. C. Gupta. J. F. Talbot. andA. Zaidi and Miss M. A. C. Jones for their assistance in obtaining thepathological material; Miss P. Little. Mr A. Buxton. Miss P. Mills.and Mr C. Day for technical assistance; and Mrs P. Kirk for typingthe manuscript. The work was supported by a grant from the trusteesto the former United Sheffield Hospitals (grant number 276).

References

I Callender GR. Malignant melanotic tumours of the eye: a studyof the histological types in 111 cases. Trans Am Acad OphthalmolOtolaryngol 1931; 36: 131-40.

2 Callender GR, Wilder HC, Ash JE. Five hundred melanomas ofthe choroid and ciliary body followed five years or longer. Am JOphthalmol 1942; 25: 962-7.

3 Flocks M. Gerende JH, Zimmerman LE. The size and shape ofmalignant melanomas of the choroid and ciliary body in relationto prognosis and histological characteristics. Trans Am AcadOphthalmol Otolaryngol 1955; 59: 740-58.

4 Paul E, Parnell BI, Fraker M. Prognosis of malignant melanomasof the choroid and ciliary body. Int Ophthalmol Clin 1962; 2:387.

5 McLean IW, Foster WD, Zimmerman LE. Prognostic factors insmall malignant melanomas of choroid and ciliary body. ArchOphthalmol 1977; 95: 48-58.

6 Gass JP. Problems in the differential diagnosis of choroidal naeviand malignant melanoma. The XXXIII Edward JacksonMemorial Lecture. Am J Ophthalmol 1977; 83: 299-323.

7 Rider CC, Taylor CB. Enolase isoenzymes in rat tissues. Electro-phoretic, chromatographic, immunological and kinetic proper-ties. Biochim Biophys Acta 1974; 365: 285-300.

8 Fletcher L, Rider CC, Taylor CB. Enolase isoenzymes III.Chromatographic and immunological characteristics of rat andbrain enolase. Biochim Biophys Acta 1976; 452: 245-52.

9 Marangos PJ, Zis AP, Clark RL, Goodwin FK. Neuronal, non-neuronal and hybrid forms of enolase in brain: structural,immunological and functional comparisons. Brain Res 1978; 150:117-33.

10 Fletcher L, Rider CC, Taylor CB, Adamson ED, Luke BM,Graham CF. Enolase isoenzymes as markers of differentiation interatocarcinoma cells and normal tissues of mouse. Dev Biol1978; 65: 462-75.

11 Schmechel DE, Brightman MW, Marangos PJ. Neurones switchfrom non-neuronal enolase to neurone-specific enolase duringdifferentiation. Brain Res 1980; 190: 195-214.

12 Schmechel DE, Marangos PJ, Brightman M. Neurone-specificenolase is a molecular marker for peripheral and central neuro-endocrine cells. Nature 1978; 276: 834-6.

247

on February 1, 2020 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.67.4.244 on 1 A

pril 1983. Dow

nloaded from

Janice A. Royds. L G. Rennie, M. A. Parsons, W. R. Timperley, and C. B. Taylor

13 Schmechel DE, Marangos PJ, Zis AP, Brightman M, GoodwinFK. Brain enolases as specific markers of neuronal and glial cells.Science 1978; 199: 313-5.

14 Teillet MA, Le Donarin N. La migration des cellules pigmentairesetudie par la methode des greffes heterospecifiques de tubenerveux chez l'ambryon d'oiseau. C R Acad Sci (D) (Paris) 1970;270: 3095-8.

15 Parsons MA, Royds JA, Taylor CB, Timperley WR. Enolaseisoenzymes as markers of cellular differentiation in normal humanfoetus and adult. J Pathol 1981; 134: 314.

16 Royds JA, Parsons MA, Rennie IG, Timperley WR, Taylor CB.

Enolase isoenzymes in benign and malignant melanocytic lesions.Diag Histopathol in press.

17 Bolton AE, HunterWM. The labelling of proteins to high specificactivities by conjugation to a '25I-containing acylating agent.Biochem J 1973; 133: 529-39.

18 Royds JA, Parsons MA, Taylor CB, Timperley WR. Enolaseisoenzyme distribution in the human brain and its tumours. JPathol 1982; 137: 37-49.

19 McLean IW, Zimmerman LE, Evans RM. Reappraisal ofCallender's spindle A type of malignant melanoma ofchoroid andciliary body. Am J Ophthalmol 1978; 86: 557-64.

248

on February 1, 2020 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.67.4.244 on 1 A

pril 1983. Dow

nloaded from